This program of research involves two projects. The goal of the first project is to explore the effects of domestic violence on children who were either victims of physical abuse, witnesses of spousal abuse, both victims and witnesses, or neither victims nor witnesses. These children were first assessed in 1988/9 when they were between 8 and 12 years of age, and the children, their parents, and their teachers have since been reinterviewed, analyses of the data are now in progress. To date, analyses have focused on demonstrating and evaluating the implications of major discrepencies between reports by different informants concerning both family violence and children's adjustment. The goal of the second project is to develop and assess techniques for enhancing the informativeness of child witnesses and for evaluating the credibility of their accounts. In one study, investigators around the United States and in Israel provided transcripts of audio- or videotaped interviews along with case files to establish case facts independent of the statements. The statements were assessed using a revision of the Criterion-Based Content Analysis (CBCA) technique, while case facts were independently rated using "ground truth" assessment scales developed by SSED staff. Accounts of events that appeared very likely to have happened indeed contained more CBCA criteria than accounts that appeared implausible, although the differences were not large. Other studies in the research program have focused on the relationship between interviewer style and the quality of information provided by young children. Several studies have confirmed that open-ended questions elicit longer and more detailed responses than more focused questions, regardless of the number of incidents experienced and the language (English or Hebrew) in which the interview was conducted. The use of anatomical dolls neither affects the association nor enhances the informativeness of young children. Recent studies have shown that interviewers can increase the length and richness of children's accounts by following SSED-designed protocols designed to probe recall memory and reduce the reliance on more focused questions which are more likely to elicit erroneous information. Currently, efforts are being made to extend and evaluate these interview protocols.